%load_ext autoreload
%autoreload 2

from lib import *
from pipeline_presets import *
from grid_search_presets import *

Data Loading, Set Splitting Functions, Data Resampling

# Load breast cancer dataset
cancer = load_breast_cancer()
cancer_df = pd.DataFrame(data=cancer.data, columns=cancer.feature_names)
cancer_df['target'] = cancer.target

# Preview
display(cancer_df.head())
display(cancer_df.describe())

	mean radius	mean texture	mean perimeter	mean area	mean smoothness	mean compactness	mean concavity	mean concave points	mean symmetry	mean fractal dimension	...	worst texture	worst perimeter	worst area	worst smoothness	worst compactness	worst concavity	worst concave points	worst symmetry	worst fractal dimension	target
0	17.99	10.38	122.80	1001.0	0.11840	0.27760	0.3001	0.14710	0.2419	0.07871	...	17.33	184.60	2019.0	0.1622	0.6656	0.7119	0.2654	0.4601	0.11890	0
1	20.57	17.77	132.90	1326.0	0.08474	0.07864	0.0869	0.07017	0.1812	0.05667	...	23.41	158.80	1956.0	0.1238	0.1866	0.2416	0.1860	0.2750	0.08902	0
2	19.69	21.25	130.00	1203.0	0.10960	0.15990	0.1974	0.12790	0.2069	0.05999	...	25.53	152.50	1709.0	0.1444	0.4245	0.4504	0.2430	0.3613	0.08758	0
3	11.42	20.38	77.58	386.1	0.14250	0.28390	0.2414	0.10520	0.2597	0.09744	...	26.50	98.87	567.7	0.2098	0.8663	0.6869	0.2575	0.6638	0.17300	0
4	20.29	14.34	135.10	1297.0	0.10030	0.13280	0.1980	0.10430	0.1809	0.05883	...	16.67	152.20	1575.0	0.1374	0.2050	0.4000	0.1625	0.2364	0.07678	0

5 rows × 31 columns

	mean radius	mean texture	mean perimeter	mean area	mean smoothness	mean compactness	mean concavity	mean concave points	mean symmetry	mean fractal dimension	...	worst texture	worst perimeter	worst area	worst smoothness	worst compactness	worst concavity	worst concave points	worst symmetry	worst fractal dimension	target
count	569.000000	569.000000	569.000000	569.000000	569.000000	569.000000	569.000000	569.000000	569.000000	569.000000	...	569.000000	569.000000	569.000000	569.000000	569.000000	569.000000	569.000000	569.000000	569.000000	569.000000
mean	14.127292	19.289649	91.969033	654.889104	0.096360	0.104341	0.088799	0.048919	0.181162	0.062798	...	25.677223	107.261213	880.583128	0.132369	0.254265	0.272188	0.114606	0.290076	0.083946	0.627417
std	3.524049	4.301036	24.298981	351.914129	0.014064	0.052813	0.079720	0.038803	0.027414	0.007060	...	6.146258	33.602542	569.356993	0.022832	0.157336	0.208624	0.065732	0.061867	0.018061	0.483918
min	6.981000	9.710000	43.790000	143.500000	0.052630	0.019380	0.000000	0.000000	0.106000	0.049960	...	12.020000	50.410000	185.200000	0.071170	0.027290	0.000000	0.000000	0.156500	0.055040	0.000000
25%	11.700000	16.170000	75.170000	420.300000	0.086370	0.064920	0.029560	0.020310	0.161900	0.057700	...	21.080000	84.110000	515.300000	0.116600	0.147200	0.114500	0.064930	0.250400	0.071460	0.000000
50%	13.370000	18.840000	86.240000	551.100000	0.095870	0.092630	0.061540	0.033500	0.179200	0.061540	...	25.410000	97.660000	686.500000	0.131300	0.211900	0.226700	0.099930	0.282200	0.080040	1.000000
75%	15.780000	21.800000	104.100000	782.700000	0.105300	0.130400	0.130700	0.074000	0.195700	0.066120	...	29.720000	125.400000	1084.000000	0.146000	0.339100	0.382900	0.161400	0.317900	0.092080	1.000000
max	28.110000	39.280000	188.500000	2501.000000	0.163400	0.345400	0.426800	0.201200	0.304000	0.097440	...	49.540000	251.200000	4254.000000	0.222600	1.058000	1.252000	0.291000	0.663800	0.207500	1.000000

8 rows × 31 columns

def get_train_test_cancer(cancer_df, test_size=0.2, random_state = global_random_state, stratify=True, resample=True) -> tuple:
    """
    Returns a tuple of (X_train, X_test, y_train, y_test) for the diabetes dataset.
    """
    
    # Split the data into features and target variable
    X = cancer_df.drop(columns=['target'])
    y = cancer_df['target']
    
    # Split the dataset into training and testing sets
    
    if stratify:
        return train_test_split(X, y, random_state=random_state, test_size=test_size, stratify=y)
    else:
        return train_test_split(X, y, random_state=random_state, test_size=0.2)

X_train, X_test, y_train, y_test = get_train_test_cancer(cancer_df, test_size=0.2, random_state=global_random_state, stratify=True, resample=True)

Exploratory Data Analysis

Pairplot (Only between means)

# Make pairplots between all features with the target variable being outcome
cancer_df_means = cancer_df[[col for col in cancer_df.columns if col.startswith("mean") or col == "target"]]
sns.pairplot(cancer_df_means, hue='target', markers='o', height=2.5, aspect=1.2, plot_kws={'s': 20})

<seaborn.axisgrid.PairGrid at 0x2b8369090>

Pairplot (Only between worsts)

# Make pairplots between all features with the target variable being outcome
cancer_df_worsts = cancer_df[[col for col in cancer_df.columns if col.startswith("worst") or col == "target"]]
sns.pairplot(cancer_df_means, hue='target', markers='o', height=2.5, aspect=1.2, plot_kws={'s': 20})

<seaborn.axisgrid.PairGrid at 0x2a4b49180>

Correlation Plot

# Make a correlation heatmap
plt.figure(figsize=(10, 8))
sns.heatmap(cancer_df.corr(), annot=True, fmt='.3f', cmap='coolwarm', square=True, cbar_kws={"shrink": .8}, annot_kws={"size": 5}, linewidths=0.5, linecolor='black')
# 

<Axes: >

Countplot

#Countplot of the target variable
plt.figure(figsize=(8, 5))
sns.countplot(x='target', data=cancer_df, palette='Set2')

/var/folders/tp/3nd20r2x6ns2txnn5kmx15dr0000gn/T/ipykernel_30431/474471373.py:3: FutureWarning: 

Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.

  sns.countplot(x='target', data=cancer_df, palette='Set2')

<Axes: xlabel='target', ylabel='count'>

Feature Distribution Plots

# Feature distribution plots
def plot_feature_distribution(df, feature):
    plt.figure(figsize=(10, 5))
    sns.histplot(df[feature], kde=True, bins=30, color='blue')
    plt.title(f'Distribution of {feature}')
    plt.xlabel(feature)
    plt.ylabel('Frequency')
    plt.grid()
    plt.show()

# Plot feature distributions for all features
for feature in cancer_df.columns[:-1]:  # Exclude the target variable 'Outcome'
    plot_feature_distribution(cancer_df, feature)

Model Fitting and Scoring

Preparation

def grid_search_preset(pipe_source, grid_search_source):
    prep_pipe = get_preprocessing_pipe(scaling=True, preprocessing="PCA")
    pipe = pipe_source(prep_pipe)
    prep_grid = get_preprocessing_grid_search(pipe)
    return grid_search_source(pipe, prep_grid, scoring='f1')

KNN

grid_search = grid_search_preset(get_knn_pipe, get_knn_grid_search)

grid_search.fit(X_train, y_train)

best_params = grid_search.best_params_
best_score = grid_search.best_score_

#Print classification report for the best model
print("KNN Report:")
y_pred = grid_search.predict(X_test)
print(classification_report(y_test, y_pred))

#Print Confusion Matrix
print("Confusion Matrix:")
print(confusion_matrix(y_test, y_pred))

# Print the best parameters and score
print("Best Parameters:", best_params)
print("F1 Score of Best Model:", best_score)

knn_grid_search = grid_search

KNN Report:
              precision    recall  f1-score   support

           0       0.87      0.93      0.90        42
           1       0.96      0.92      0.94        72

    accuracy                           0.92       114
   macro avg       0.91      0.92      0.92       114
weighted avg       0.92      0.92      0.92       114

Confusion Matrix:
[[39  3]
 [ 6 66]]
Best Parameters: {'KNN__algorithm': 'auto', 'KNN__n_neighbors': 7, 'KNN__weights': 'uniform', 'PCA__n_components': 5, 'PCA__random_state': 42, 'Scaler': StandardScaler()}
F1 Score of Best Model: 0.9771980285254622

Naive Bayes

grid_search = grid_search_preset(get_naive_bayes_pipe, get_naive_bayes_grid_search)

grid_search.fit(X_train, y_train)

best_params = grid_search.best_params_
best_score = grid_search.best_score_

# Print classification report for the best model
print("Naive Bayes Report:")
y_pred = grid_search.predict(X_test)
print(classification_report(y_test, y_pred))

print("Confusion Matrix:")
print(confusion_matrix(y_test, y_pred))

print("Best Parameters:", best_params)
print("F1 Score of Best Model:", best_score)

naive_bayes_grid_search = grid_search

Naive Bayes Report:
              precision    recall  f1-score   support

           0       0.84      0.88      0.86        42
           1       0.93      0.90      0.92        72

    accuracy                           0.89       114
   macro avg       0.88      0.89      0.89       114
weighted avg       0.90      0.89      0.90       114

Confusion Matrix:
[[37  5]
 [ 7 65]]
Best Parameters: {'PCA__n_components': 5, 'PCA__random_state': 42, 'Scaler': StandardScaler(), 'clf': GaussianNB(), 'clf__var_smoothing': 1e-09}
F1 Score of Best Model: 0.9547040124626331

Logistic Regression

# "UserWarning: Setting penalty='None' will ignore the C and l1_ratio parameters"
# Repeated warnings coming from logreg grid search can be suppressed with this file 
# Penalty=None is not supported for the liblinear solver, so 1/4 fail, but that's okay.
from remove_warnings import *

with suppress_stdout_stderr():
    warnings.filterwarnings('ignore', category=UserWarning)
    
    grid_search = grid_search_preset(get_log_reg_pipe, get_log_reg_grid_search)
    
    grid_search.fit(X_train, y_train)
    
best_params = grid_search.best_params_
best_score = grid_search.best_score_

# Print classification report for the best model
print("Logistic Regression Report:")
y_pred = grid_search.predict(X_test)
print(classification_report(y_test, y_pred))

print("Confusion Matrix:")
print(confusion_matrix(y_test, y_pred))

print("Best Parameters:", best_params)
print("F1 Score of Best Model:", best_score)

log_reg_grid_search = grid_search

Logistic Regression Report:
              precision    recall  f1-score   support

           0       0.91      0.95      0.93        42
           1       0.97      0.94      0.96        72

    accuracy                           0.95       114
   macro avg       0.94      0.95      0.94       114
weighted avg       0.95      0.95      0.95       114

Confusion Matrix:
[[40  2]
 [ 4 68]]
Best Parameters: {'LogisticRegression__C': 1, 'LogisticRegression__penalty': 'l2', 'LogisticRegression__solver': 'liblinear', 'PCA__n_components': 5, 'PCA__random_state': 42, 'Scaler': StandardScaler()}
F1 Score of Best Model: 0.98068486433475

Decision Trees

grid_search = grid_search_preset(get_decision_tree_pipe, get_decision_tree_grid_search)

grid_search.fit(X_train, y_train)

best_params = grid_search.best_params_
best_score = grid_search.best_score_

# Print classification report for the best model
print("Decision Tree Report:")
y_pred = grid_search.predict(X_test)
print(classification_report(y_test, y_pred))

print("Confusion Matrix:")
print(confusion_matrix(y_test, y_pred))

print("Best Parameters:", best_params)
print("F1 Score of Best Model:", best_score)

decision_tree_grid_search = grid_search

Decision Tree Report:
              precision    recall  f1-score   support

           0       0.84      0.98      0.90        42
           1       0.98      0.89      0.93        72

    accuracy                           0.92       114
   macro avg       0.91      0.93      0.92       114
weighted avg       0.93      0.92      0.92       114

Confusion Matrix:
[[41  1]
 [ 8 64]]
Best Parameters: {'DecisionTree__criterion': 'gini', 'DecisionTree__max_depth': 5, 'DecisionTree__min_samples_leaf': 1, 'DecisionTree__min_samples_split': 5, 'PCA__n_components': 3, 'PCA__random_state': 42, 'Scaler': StandardScaler()}
F1 Score of Best Model: 0.9609786791365739

Support Vector Machines

Linear Support Vector Machines (Faster: able to utilize larger grid search space)

grid_search = grid_search_preset(get_lin_svm_pipe, get_lin_svm_grid_search)

grid_search.fit(X_train, y_train)

best_params = grid_search.best_params_
best_score = grid_search.best_score_

# Print classification report for the best model
print("Linear SVM Report:")
y_pred = grid_search.predict(X_test)
print(classification_report(y_test, y_pred))

print("Confusion Matrix:")
print(confusion_matrix(y_test, y_pred))

print("Best Parameters:", best_params)
print("F1 Score of Best Model:", best_score)

linear_svm_grid_search = grid_search

Linear SVM Report:
              precision    recall  f1-score   support

           0       0.91      0.95      0.93        42
           1       0.97      0.94      0.96        72

    accuracy                           0.95       114
   macro avg       0.94      0.95      0.94       114
weighted avg       0.95      0.95      0.95       114

Confusion Matrix:
[[40  2]
 [ 4 68]]
Best Parameters: {'PCA__n_components': 5, 'PCA__random_state': 42, 'SVM__C': 0.1, 'SVM__penalty': 'l1', 'Scaler': StandardScaler()}
F1 Score of Best Model: 0.9788528506722752

Nonlinear / Kernel Support Vector Machines

grid_search = grid_search_preset(get_kernel_svm_pipe, get_kernel_svm_grid_search)

grid_search.fit(X_train, y_train)

best_params = grid_search.best_params_
best_score = grid_search.best_score_

print("Kernel SVM Report:")
y_pred = grid_search.predict(X_test)
print(classification_report(y_test, y_pred))

print("Confusion Matrix:")
print(confusion_matrix(y_test, y_pred))

print("Best Parameters:", best_params)
print("F1 Score of Best Model:", best_score)

kernel_svm_grid_search = grid_search

Kernel SVM Report:
              precision    recall  f1-score   support

           0       0.91      0.95      0.93        42
           1       0.97      0.94      0.96        72

    accuracy                           0.95       114
   macro avg       0.94      0.95      0.94       114
weighted avg       0.95      0.95      0.95       114

Confusion Matrix:
[[40  2]
 [ 4 68]]
Best Parameters: {'PCA__n_components': 5, 'PCA__random_state': 42, 'SVM__C': 10, 'SVM__gamma': 'scale', 'SVM__kernel': 'rbf', 'SVM__probability': False, 'Scaler': StandardScaler()}
F1 Score of Best Model: 0.9806508566197307

Final Score Comparison

# Final score comparison
print("\n===Final F1 Score Comparison===\n")
print(f"KNN:{round(knn_grid_search.best_score_ * 100, 2)}%")
print(f"Naive Bayes:{round(naive_bayes_grid_search.best_score_ * 100, 2)}%")
print(f"Logistic Regression:{round(log_reg_grid_search.best_score_ * 100, 2)}%")
print(f"Decision Tree:{round(decision_tree_grid_search.best_score_ * 100, 2)}%")
print(f"Linear SVM:{round(linear_svm_grid_search.best_score_ * 100, 2)}%")
print(f"Kernel SVM:{round(kernel_svm_grid_search.best_score_ * 100, 2)}%")

===Final F1 Score Comparison===

KNN:97.72%
Naive Bayes:95.47%
Logistic Regression:98.07%
Decision Tree:96.1%
Linear SVM:97.89%
Kernel SVM:98.07%

Time Reports for all training

print("=========KNN==========")
print_time_report(knn_grid_search)
print("=========Naive Bayes==========")
print_time_report(naive_bayes_grid_search)
print("=========Logistic Regression==========")
print_time_report(log_reg_grid_search)
print("=========Decision Tree==========")
print_time_report(decision_tree_grid_search)
print("=========Linear SVM==========")
print_time_report(linear_svm_grid_search)
print("=========Kernel SVM==========")
print_time_report(kernel_svm_grid_search)

=========KNN==========

=== Training Time by 'Scaler' ===
param_Scaler
StandardScaler()    0.012959
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__n_components' ===
param_PCA__n_components
4    0.010859
2    0.011056
3    0.013773
5    0.015075
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__random_state' ===
param_PCA__random_state
42    0.012691
Name: mean_fit_time, dtype: float64

=== Training Time by 'KNN__n_neighbors' ===
param_KNN__n_neighbors
3    0.009548
5    0.009905
9    0.014441
7    0.016870
Name: mean_fit_time, dtype: float64

=== Training Time by 'KNN__weights' ===
param_KNN__weights
distance    0.011920
uniform     0.013462
Name: mean_fit_time, dtype: float64

=== Training Time by 'KNN__algorithm' ===
param_KNN__algorithm
kd_tree      0.009392
ball_tree    0.010237
brute        0.010730
auto         0.020405
Name: mean_fit_time, dtype: float64
=========Naive Bayes==========

====== Training Time by Group 1 ======

=== Training Time by 'Scaler' ===
param_Scaler
StandardScaler()    0.008058
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__n_components' ===
param_PCA__n_components
2    0.005977
5    0.006800
4    0.008677
3    0.008859
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__random_state' ===
param_PCA__random_state
42    0.007578
Name: mean_fit_time, dtype: float64

=== Training Time by 'clf' ===
param_clf
GaussianNB()    0.007578
Name: mean_fit_time, dtype: float64

=== Training Time by 'clf__var_smoothing' ===
param_clf__var_smoothing
1.000000e-09    0.007111
1.000000e-07    0.007402
1.000000e-08    0.008222
Name: mean_fit_time, dtype: float64

====== Training Time by Group 2 ======

=== Training Time by 'Scaler' ===
param_Scaler
StandardScaler()    0.009551
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__n_components' ===
param_PCA__n_components
5    0.007317
4    0.007401
2    0.009341
3    0.009866
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__random_state' ===
param_PCA__random_state
42    0.008481
Name: mean_fit_time, dtype: float64

=== Training Time by 'clf' ===
param_clf
BernoulliNB()    0.008481
Name: mean_fit_time, dtype: float64

=== Training Time by 'clf__alpha' ===
param_clf__alpha
0.5    0.008071
1.0    0.008891
Name: mean_fit_time, dtype: float64

=== Training Time by 'clf__binarize' ===
param_clf__binarize
0.0    0.008322
0.5    0.008640
Name: mean_fit_time, dtype: float64
=========Logistic Regression==========

=== Training Time by 'Scaler' ===
param_Scaler
StandardScaler()    0.004787
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__n_components' ===
param_PCA__n_components
2    0.004633
3    0.004845
4    0.005006
5    0.005215
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__random_state' ===
param_PCA__random_state
42    0.004925
Name: mean_fit_time, dtype: float64

=== Training Time by 'LogisticRegression__C' ===
param_LogisticRegression__C
0.01      0.004727
10.00     0.004782
100.00    0.004839
0.10      0.005043
1.00      0.005231
Name: mean_fit_time, dtype: float64

=== Training Time by 'LogisticRegression__penalty' ===
param_LogisticRegression__penalty
l2    0.004878
Name: mean_fit_time, dtype: float64

=== Training Time by 'LogisticRegression__solver' ===
param_LogisticRegression__solver
liblinear    0.003953
lbfgs        0.005896
Name: mean_fit_time, dtype: float64
=========Decision Tree==========

=== Training Time by 'Scaler' ===
param_Scaler
StandardScaler()    0.017469
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__n_components' ===
param_PCA__n_components
2    0.014665
3    0.015257
4    0.017284
5    0.017388
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__random_state' ===
param_PCA__random_state
42    0.016148
Name: mean_fit_time, dtype: float64

=== Training Time by 'DecisionTree__criterion' ===
param_DecisionTree__criterion
entropy    0.015955
gini       0.016342
Name: mean_fit_time, dtype: float64

=== Training Time by 'DecisionTree__max_depth' ===
param_DecisionTree__max_depth
15    0.013242
10    0.016003
5     0.017335
Name: mean_fit_time, dtype: float64

=== Training Time by 'DecisionTree__min_samples_split' ===
param_DecisionTree__min_samples_split
2     0.015987
10    0.016134
5     0.016324
Name: mean_fit_time, dtype: float64

=== Training Time by 'DecisionTree__min_samples_leaf' ===
param_DecisionTree__min_samples_leaf
2    0.015004
1    0.015848
4    0.017592
Name: mean_fit_time, dtype: float64
=========Linear SVM==========

=== Training Time by 'Scaler' ===
param_Scaler
StandardScaler()    0.011021
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__n_components' ===
param_PCA__n_components
2    0.008914
3    0.010188
5    0.010468
4    0.011788
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__random_state' ===
param_PCA__random_state
42    0.01034
Name: mean_fit_time, dtype: float64

=== Training Time by 'SVM__C' ===
param_SVM__C
1.000      0.009771
10.000     0.009981
0.001      0.010090
0.100      0.010398
50.000     0.010508
0.010      0.010739
100.000    0.010890
Name: mean_fit_time, dtype: float64

=== Training Time by 'SVM__penalty' ===
param_SVM__penalty
l1    0.010072
l2    0.010607
Name: mean_fit_time, dtype: float64
=========Kernel SVM==========

====== Training Time by Group 1 ======

=== Training Time by 'Scaler' ===
param_Scaler
StandardScaler()    0.013573
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__n_components' ===
param_PCA__n_components
4    0.025761
5    0.026543
3    0.030225
2    0.034988
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__random_state' ===
param_PCA__random_state
42    0.029379
Name: mean_fit_time, dtype: float64

=== Training Time by 'SVM__C' ===
param_SVM__C
100.00    0.009573
50.00     0.009957
10.00     0.010370
5.00      0.010946
0.05      0.013030
0.01      0.015324
0.10      0.019631
1.00      0.079760
Name: mean_fit_time, dtype: float64

=== Training Time by 'SVM__kernel' ===
param_SVM__kernel
sigmoid    0.011195
rbf        0.012176
poly       0.012911
linear     0.189618
Name: mean_fit_time, dtype: float64

=== Training Time by 'SVM__probability' ===
param_SVM__probability
False    0.029379
Name: mean_fit_time, dtype: float64

=== Training Time by 'SVM__gamma' ===
param_SVM__gamma
0.0      0.012911
scale    0.039780
Name: mean_fit_time, dtype: float64

====== Training Time by Group 2 ======

=== Training Time by 'Scaler' ===
param_Scaler
StandardScaler()    0.013573
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__n_components' ===
param_PCA__n_components
4    0.025761
5    0.026543
3    0.030225
2    0.034988
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__random_state' ===
param_PCA__random_state
42    0.029379
Name: mean_fit_time, dtype: float64

=== Training Time by 'SVM__C' ===
param_SVM__C
100.00    0.009573
50.00     0.009957
10.00     0.010370
5.00      0.010946
0.05      0.013030
0.01      0.015324
0.10      0.019631
1.00      0.079760
Name: mean_fit_time, dtype: float64

=== Training Time by 'SVM__kernel' ===
param_SVM__kernel
sigmoid    0.011195
rbf        0.012176
poly       0.012911
linear     0.189618
Name: mean_fit_time, dtype: float64

=== Training Time by 'SVM__probability' ===
param_SVM__probability
False    0.029379
Name: mean_fit_time, dtype: float64

=== Training Time by 'SVM__gamma' ===
param_SVM__gamma
0.0      0.012911
scale    0.039780
Name: mean_fit_time, dtype: float64

====== Training Time by Group 3 ======

=== Training Time by 'Scaler' ===
param_Scaler
StandardScaler()    0.013828
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__n_components' ===
param_PCA__n_components
3    0.011362
2    0.011795
5    0.013989
4    0.014499
Name: mean_fit_time, dtype: float64

=== Training Time by 'PCA__random_state' ===
param_PCA__random_state
42    0.012911
Name: mean_fit_time, dtype: float64

=== Training Time by 'SVM__C' ===
param_SVM__C
0.10    0.012138
0.01    0.013002
1.00    0.013595
Name: mean_fit_time, dtype: float64

=== Training Time by 'SVM__kernel' ===
param_SVM__kernel
poly    0.012911
Name: mean_fit_time, dtype: float64

=== Training Time by 'SVM__degree' ===
param_SVM__degree
3.0    0.012437
2.0    0.012645
4.0    0.012986
5.0    0.013577
Name: mean_fit_time, dtype: float64

=== Training Time by 'SVM__probability' ===
param_SVM__probability
False    0.012911
Name: mean_fit_time, dtype: float64

=== Training Time by 'SVM__gamma' ===
param_SVM__gamma
1.000000e-09    0.012911
Name: mean_fit_time, dtype: float64